Take the financials of a laghubitta company and a commercial bank, strip the names off, and hand them to most retail investors. They will read both the same way. Net profit up, EPS up, book value up, so the stock is a buy. That instinct is wrong for microfinance, and it is exactly how people get trapped.
The reason is simple. A microfinance balance sheet in Nepal is shaped by rules a bank never faces, and the numbers that actually signal trouble are not the ones on the front page. If you want to read a microfinance balance sheet, PAR is where Nepal investors should start, not the profit line. Portfolio at Risk tells you what the loan book is really worth. The regulated interest spread tells you how much the company is even allowed to earn on it. Miss either one and you are guessing.
This is the sector-specific health check. It assumes you already know the general ratio toolkit for banks. If you do not, read our piece on banking ratios (NPL, CD and CAR) first. We will not re-explain those here. What we will do is show where a laghubitta lender differs, and what deterioration looks like before it hits the headline number.
Why the usual bank read fails here
Start with the business itself. A commercial bank lends against collateral, funds itself from a broad deposit base, and earns on scale. A microfinance company (laghubitta bittiya sanstha) lends small, mostly uncollateralized, mostly to group-guaranteed borrowers in rural and semi-urban Nepal, and funds a large share of that book with wholesale borrowing from commercial banks rather than public deposits. The whole cost and risk structure is different. We cover how that model actually works, and why it is fragile, in the microfinance business model in Nepal.
The practical consequence for reading financials is this. A bank’s asset quality shows up slowly, cushioned by collateral you can recover. A microfinance book has almost no collateral to fall back on, so once repayment discipline breaks in a village group, the loss is close to the full loan. Deterioration is faster and sharper. And because the profit margin is capped by the regulator, the company cannot simply reprice risk upward to compensate. So the two variables that dominate everything are the quality of the loan book and the spread the company is allowed to keep. Everything else is secondary.
PAR: the number that actually tells you the truth
Portfolio at Risk is the share of the total loan book that has a payment overdue past a set number of days, usually reported as PAR>30 (more than 30 days late) and PAR>90 (more than 90 days late). It is the microfinance industry’s standard asset-quality metric, and it exists because microfinance loans are short-term and repaid in small, frequent installments. A missed installment is an early warning, so the sector watches a 30-day window rather than waiting for the longer classification cycle a corporate loan would use.
Here is the key thing. PAR counts the entire outstanding balance of a loan the moment it goes overdue, not just the missed installment. If a borrower owes NPR 80,000 and misses one weekly payment, the whole NPR 80,000 sits in the PAR bucket. That is why PAR moves fast and why it flatters nobody. It is the truest single read on whether the book is healthy.
In Nepal, the regulatory cousin of PAR is the non-performing loan (NPL) ratio, which Nepal Rastra Bank (NRB) reporting uses, and the two track each other closely. The sector-level trend has been ugly. Industry NPLs rose from about 2.6 percent in mid-July 2022 to roughly 7.2 percent by mid-April 2025. That near-tripling is the single most important fact about the sector over the last three years, and it barely shows up if you only read net profit.
Individual companies are worse than the average. Ratings and disclosures have shown gross NPLs at some laghubitta companies in the mid-teens, for example one issuer moving from around 12.4 percent (mid-July 2023) to about 16.5 percent (mid-January 2024), while the best-run names have sat below 1 percent. The spread between the best and worst operators is enormous, which is the whole point. The sector average hides everything.
What does deteriorating look like on PAR? Watch three things. First, the direction and speed, not the level. PAR climbing from 3 to 5 to 8 percent across three quarters is a screaming signal even if 8 percent still sounds survivable. Second, the gap between PAR>30 and PAR>90. If PAR>90 is catching up to PAR>30, early delinquency is hardening into permanent loss rather than curing. Third, PAR relative to loan-loss provisions. If the book is going bad faster than provisions are being built, reported profit is borrowing from a future writeoff.
The spread is capped, so the profit engine is regulated
Now the part that makes microfinance unlike almost any other listed sector in Nepal. The interest margin, the spread between what a laghubitta pays for funds and what it charges borrowers, is not set by the market. It is set, or at least ceilinged, by NRB.
The rules have moved, so read the year on any figure carefully. For years NRB effectively capped the microfinance lending rate and required a maximum interest spread (historically limited to 7 percent, with a sub-limit on how much administrative cost could be loaded in). A flat 15 percent maximum lending rate was also enforced. Then from Shrawan 1, 2082 (July 16, 2025), NRB moved microfinance onto a base-rate framework: for new loans, the rate is the company’s own base rate plus a premium of at most 3 percentage points, recalculated quarterly, with an overall ceiling of the latest commercial-bank average base rate plus 9 percentage points. Loans booked before that date remained subject to the 15 percent ceiling.
Forget the exact numbers for a moment and hold the structural point, because that is the evergreen part. The profit engine of a microfinance company runs inside a box the regulator draws. Management cannot widen the spread to earn its way out of trouble the way an unregulated lender could. When funding costs rise, and microfinance funding is largely wholesale borrowing from banks that reprices with the rate cycle, the lending rate cannot rise freely to match. The spread gets squeezed from the funding side while the ceiling holds the top down. That is why NRB monetary policy matters so directly to this sector; we cover the transmission in how NRB monetary policy moves NEPSE.
So when you read a microfinance income statement, do not just look at net interest income. Look at the realized spread and compare it to the regulatory ceiling. A company already charging near the cap has no room left. If its funding cost then rises, the margin compresses, and there is nothing management can legally do about it. That single constraint tells you more about near-term earnings than any growth story in the notes.
Self-sufficiency: can it cover its own costs
Because the spread is capped and the cost of serving thousands of tiny rural loans is high, the question of whether a laghubitta can actually cover its own running costs is not academic. Two ratios answer it.
Operational self-sufficiency (OSS) is operating income divided by the sum of operating expenses, financial (funding) costs, and loan-loss provisions. Above 100 percent means the company covers its full operating cost from its own revenue. Below 100 percent means it is bleeding. Financial self-sufficiency (FSS) is the stricter version, adjusting for subsidized funding and inflation to test whether the operation stands on genuinely commercial terms.
For a listed, profit-oriented laghubitta, these are usually comfortably above 100 percent, so the more useful reading is the trend. OSS drifting down toward 100 tells you the cost structure is eating the capped margin. And here is where it links back to PAR. Loan-loss provisions sit in the OSS denominator. So a rising PAR feeds straight into deteriorating self-sufficiency: bad loans force higher provisions, higher provisions push OSS down, and a company that looked sustainable a year ago starts failing to cover itself. PAR is upstream of everything.
Concentration and the multiple-borrowing problem
A bank worries about a few large corporate exposures. A microfinance company faces the opposite risk with the same name: concentration by geography, by sector (much of the book is agriculture, livestock and petty trade), and by borrower behavior across the whole industry.
The behavioral risk is specific to Nepal, and it is serious. Multiple borrowing, one household taking loans from several laghubitta at once, inflated the sector during the boom and then blew up in the 2023 anti-microfinance protests and loan-repayment agitation. An NRB committee found some borrowers juggling loans from more than 24 institutions. NRB responded by capping how many lenders a borrower may use (restricted to a single MFI from around November 2022, later eased to two by mid-2024) and cutting per-borrower loan limits (roughly NPR 5 lakh under group guarantee and NPR 7 lakh against collateral, down from an earlier NPR 15 lakh).
Read this off the financials by asking where the growth came from. A book that ballooned in 2079 and 2080 (2022 and 2023) was very likely inflated by the same borrowers appearing on several lenders’ books. When the rules tightened, that phantom growth reversed, disbursement fell, and the loans already out turned bad. That is precisely the mechanism behind the sector NPL spike. So a laghubitta showing rapid book growth in that window is not a strength to reward; it is a flag to interrogate.
The classic trap: growth up, PAR quietly up
Now put it together, because this is the whole reason to read a laghubitta differently.
Picture a company that grew its loan book 30 percent last year. EPS is up. The board is proposing a healthy dividend. On a bank-style read, it looks like a winner. But its PAR>30 drifted from 4 to 7 percent over the same year, PAR>90 is rising faster than PAR>30, provisions have not kept pace, and it is already lending at the regulatory ceiling with no spread room left. That is not a growth stock. That is a company booking today’s interest income on loans that are quietly going bad, funded by borrowers who may be over-leveraged across the industry, with no legal ability to widen its margin when funding costs rise.
Growing the book while PAR quietly rises is the classic microfinance trap. The growth flatters the near-term profit, and the deteriorating asset quality is the bill that arrives later, as writeoffs, as forced provisioning, as a dividend that suddenly cannot be paid. And note the distinction that catches investors out here more than in any other sector: reported net profit and the cash a company can actually distribute are not the same thing once provisioning and regulatory reserves take their cut. We unpack that gap in distributable profit versus net profit.
The verdict
Read a microfinance balance sheet in this order, not the front-page order. PAR first, and its trend and its PAR>90 tail, because that is the true state of the assets. Then the realized spread against the regulatory ceiling, because that caps what the company can earn on those assets no matter how well it is run. Then OSS, to check the capped margin still covers the cost of doing business. Then the growth story, treated with suspicion, especially any surge during the multiple-borrowing boom years.
Do that and two numbers, PAR and the capped spread, will tell you almost everything about a microfinance company’s near-term earnings and risk. The profit line is the last thing to trust, not the first. A laghubitta that is growing its book while PAR climbs and its margin sits pinned against the cap is not compounding. It is running down a road the regulator already narrowed, and the reckoning shows up on the balance sheet a year after it showed up in the villages.
This is analysis, not financial advice.
